Using logic to decide whether or not to bring in (a sample) data.

Question

i have a report that works like this.
Its broken into 2 different queries that are joined by data links. The First query grabs the most recent data (using max date) based on the group by function at the bottom of the query.
select max(ops$penlims.get_colldate(b.sample_Id)) fdate,
b.user_sampleid fuid, a.method fmethod, ops$Penlims.get_collpt_consult(b.sample_id) as ftype
from nais_tasks a, nais_samples b
where a.sample_id = b.sample_id
and ops$penlims.get_colldate(b.sample_id) between :inStartDate AND :inEndDate
and b.user_sampleid like :well
and ops$penlims.get_collpt_consult(b.sample_id) like :pType
and ops$penlims.get_collpt_consult(b.sample_id) not like '%EFF%'
and a.condition in ('APPROVED')
and not a.method in ('BALANCE PROGRAM','528 (VERSION 1)')
and a.method not like '300%ANALY%ER%'
group by b.user_sampleid, ops$penlims.get_colldate(b.sample_Id), a.method, ops$Penlims.get_collpt_consult(b.sample_id)
order by user_sampleid, ops$penlims.get_colldate(b.sample_Id), a.method, ops$penlims.get_collpt_consult(b.sample_id)
The second query is joined by a data link to the first query and is primarily just the part of the report that brings in the data to populate the report.
Here is the second query:
select distinct a.sample_id SSID, a.user_sampleid SUID, c.number_value, b.operation,
ops$penlims.get_colldate(a.sample_id) SDATE, ops$penlims.get_collpt_consult(a.sample_id) STYPE, b.method SMETHOD,
b.task_id STID, c.units sunits, b.condition TCond, c.condition RCond,
ops$penlims.get_colltime(a.sample_id) STIME,
c.component,
ops$penlims.get_conc(result_id) gconc,
ops$penlims.get_collpt_consult(a.sample_id) gcollpt
from nais_results c, nais_tasks b, nais_samples a
where a.sample_id = b.sample_id
and b.task_id = c.task_id
AND ops$penlims.get_colldate(a.sample_id) between :inStartDate AND :inEndDate
and user_sampleid like :well
and ops$penlims.get_collpt_consult(a.sample_id) LIKE :pType
and c.condition in ('APPROVED')
and b.condition in ('APPROVED')
and ops$penlims.get_collpt(a.sample_id) LIKE '%PROD%'
and lower (component) not like '%std%'
My problem is this. My boss is happy with the report in that it brings in the most recent (ops$penlim.get_collpt_consult) sample_type for each (user_sampleid) well and groups/orders them correctly. Today he asked that if he passed in a parameter such as %RAW% into :ptype and the report returns RAW/HOP and STATION RAW for different dates; both the max date for the repective sample_type, is there a way to take the RAW/HOP sample over the STATION RAW if they both exist for that particular well (user_sampleid) or leave the STATION RAW sample in the report if no RAW/HOP sample exists. I can clarify any questions and could probably write a cursor to do this. Is there a faster way than writing a cursor? maybe a slight logic or program unit mod i can do?
Thanks for even reading this far and thanks for ANY input!!
Mike Mullowney

Answer 1

I just was reading a thread about dpi on the DW forum and
also speaking from experience they had a long thread and several
posts conducted tests proving that dpi does not really make a big
difference whether it's 72 or 300 dpi.
dpi dots per inch come into play, with print generally. Some
argued that 72 verses 300 used up more bandwidth but I did not see
any significant difference when I tested the same graphic pixel
size with one at 72 dpi and the other 300 dpi the amount of pixels
in an image ppi is what matters for monitor display.
For me most situations dynamic or embedded, I determine which
way to go by how many pictures I need to show.
Another solution I have used is to create your slide show
using Flash video which will buffer and be able to use many photos.
Hope any of this helps.
INVT